ratio (FCR; OXPHOS/ETS) of 0.87 (Table 1 and
table S1), indicating that phosphorylation ca-
pacity limits maximum OXPHOS respiration
( 13 ). The LEAK respiratory capacity in sea
otter skeletal muscle resulted in a coupling
control ratio (CCR; LEAK/OXPHOS) of 0.41
(Table 1 and table S1), indicating that leak
respiratory capacity accounts for up to 41% of
OXPHOS capacity. This mitochondrial LEAK
capacity is elevated compared with that of
other mammals (Table 1); absolute LEAK ca-
pacity and CCR are, respectively, two and
four times that of comparably sized Alaskan
husky Iditarod dogs. These values are also
higher than comparable measures reported
for any known mammal, excluding extremely
small mice species with body mass ranging
from 5 to 22 g ( 14 ). Increased thermogenic
mitochondrial leak is predicted for polar ani-
mals relying on metabolic heat production to
maintain body temperature ( 6 ) and is described
here in sea otter skeletal muscle.
Because of the scaling of surface area to
body mass, smaller neonatal mammals are
particularly susceptible to heat loss ( 4 , 15 ). In
addition, metabolically active skeletal muscle
constitutes a smaller relative portion of total
body mass in younger, less physically devel-
oped individuals ( 16 ), and their immature
skeletal muscle tends to be less thermogenic
( 17 ). Sea otters are altricial; born relatively
inactive, they depend on their mothers for
feeding and grooming and become progres-
sively more active as they mature during the
first 3 months of life ( 18 , 19 ). Given the broad
ranges of body mass and age classes included
in our data, we determined how sea otter
skeletal muscle respiratory capacity changed
throughout ontogenetic development.
Respiratory capacity was surprisingly con-
sistent across all age classes, and Pearson cor-
relation indicated that body mass was not
correlated with respiratory measures of LEAK
(r= 0.172;P=0.457;Fig.1A),OXPHOS(r=
0.274;P=0.229;Fig.1A),orETS(r= 0.294;
P= 0.197). Body mass was likewise not pre-
dictive of FCR (r=−0.112;P= 0.628) or CCR
(r=−0.220;P= 0.339) (Fig. 1B). In altricial
mammals, skeletal muscle is underdeveloped
at birth and is delayed in developing adult
measures of mechanical function ( 20 ) as
well as metabolic and thermogenic capac-
ity ( 21 ). However, sea otter muscle metabolic
capacity appears fully developed in even the
smallest neonates, although other measures
of muscle maturity, including muscle mass
and myoglobin concentration, do not reach
adult levels until animals are ~2 years of age
( 16 ). The consistent skeletal muscle mitochon-
drial leak capacity and CCR across a broad
range of body mass and age classes indicate
that mitochondrial leak does not constitute
a greater portion of total OXPHOS capacity
in smaller, less mature individuals as might
be predicted.
Samples collected from two animals in un-
usual circumstances provided additional insights.
Included in our study was one captive-raised,
rehabilitated juvenile. This individual was housed
using a flow-through system of water from its
natural habitat, ensuring an endemic temper-
ature range. Despite reduced functional demand
on skeletal muscle for foraging, diving, and
swimming, this sea otter was metabolically
indistinguishable from its wild counterparts
(Fig. 1, A and B; blue). This is consistent with
a skeletal muscle metabolic capacity that is
modulated by thermogenic demand and not
by contractile muscle load. Other thermally
challenged marine mammals, including Weddell
seals (Leptonychotes weddellii)( 22 ) and northern
elephant seals (Mirounga angustirostris)( 23 ),also demonstrate similarly high neonatal muscle
metabolic capacity ( 5 ).
One stranded, emaciated adult was an excep-
tion to the consistent high metabolic capacity
seen in this study. Owing to its extensive den-
tal wear, this individual was estimated to be
>12 years of age; it was dehydrated and ema-
ciated at the time of stranding, which suggests
that it may have had difficulty feeding. This
geriatric sea otter demonstrated a universally
profound reduction in skeletal muscle metabolic
capacity, with OXPHOS and LEAK respiratory
capacity only 34% and 36% of the average value,
respectively (Fig. 1A, red). Despite the metabolic
disturbance, LEAK, OXPHOS, and ETS respi-
ratory capacities were uniformly reduced and
respiratory ratios of FCR and CCR remained
constant (Fig. 1B, red). Although the cause of the
emaciation and stranding cannot be deduced,
metabolic factors may have affected thermo-
genic capacity and contributed to stranding.
To estimate the potential metabolic contri-
bution of skeletal muscle to sea otter resting
metabolic rate and thermogenesis throughout
life, we combined our measures of skeletal
muscle leak capacity with published measures
of scaled sea otter muscle mass to determine
whole-body skeletal muscle leak capacity (Fig. 2;
see supplementary methods for calculations). In
seaotterswithbodymassgreaterthan~9kg,
skeletal muscle leak capacity exceeds whole-body
resting oxygen consumption. However, at less
than ~9 kg, skeletal muscle leak is not indepen-
dently adequate to fully account for the resting
metabolic rate owing to the lower relative muscle
mass. For smaller animals to maintain body
temperature, maximal use of thermogenic leak
from skeletal muscle and other tissues may
be critical. Smaller animals may also require
additional muscle thermogenesis via either
shivering or sarcoplasmic reticulum calcium
leak to use a greater portion of skeletal muscle
respiratory capacity (fig. S1). Notably, neonatal
pups have dense natal fur and spend most of
their time resting on the mother’s abdomen,
so there is little heat loss to water until the age
of 1 month ( 24 ).
Thermogenic leak metabolic capacity in sea
otter muscle is elevated across all age and size
classes and, if extrapolated to whole-body mus-
cle mass, is adequate to explain the elevated
resting metabolic rate of adult sea otters. How-
ever, smaller, less mature individuals with lower
relative muscle mass require either non-leak
muscle thermogenesis (e.g., shivering, active
grooming) or other metabolically active tissues
to account for their elevated resting metabo-
lism. Although mass-specific LEAK respiratory
capacity does not change throughout ontoge-
netic development in sea otter skeletal muscle
(Fig.1A),skeletalmuscleleakcapacityconstitutes
a greater portion of resting metabolic rate in
larger animals owing to larger relative muscle
mass (Fig. 2).224 9JULY2021•VOL 373 ISSUE 6551 sciencemag.org SCIENCE
Fig. 2. Whole-body estimates of
sea otter metabolism scaled to
body mass.Estimated scaled
whole-body metabolic rates include
predicted [BMR; predicted on the
basis of scaled rate for eutherian
mammals ( 3 )], sea otter resting
metabolic rate, and whole-body
skeletal muscle leak capacity.
RESEARCH | REPORTS